Newsletter

Recommendations

Info

Research • Center of Excellence SAFE • Quarter 1/2014 Global Financial Turmoil: Does Connectedness Matter? Loriana Pelizzon Goethe University & SAFE Recent crises put the spotlight on systemic risk. Despite the absence of a widely accepted definition of this term and despite most regulators and policy makers being convinced that systemic events can be only identified after the fact, we believe that measuring the degree of connectedness among financial institutions and monitoring its evolution can provide meaningful information on where systemic risk is coming from and when a crisis is approaching. During the last few decades, the global market has faced a series of major financial crises: the 1997 Asian crisis, the 2008/09 financial crisis and the recent European banking and sovereign debt crisis. These heterogeneous crises all have something in common: they originated in specific countries and specific financial sectors, spreading over to other geographical areas and financial markets, and affecting the real economy through a severe spillover effect. According to these characteristics, the crises mentioned can be defined as “systemic”. Hence, we question whether the connectedness of institutions operating in the financial markets played a role and whether the evolution of such connectedness could be an indicator of an approaching crisis. To answer this question, my coauthors and I focused on different players in the financial markets, applying several econometric measures with two common denominators: Granger causality tests and the usage of market data instead of accounting data. Our research allowed us to infer a substantial robust relationship between the degree of connectedness and the market situation. This would suggest that connectedness measures based on a Granger causality test are potentially good indicators for detecting systemically relevant institutions and for providing an early warning of systemic crises. “On a new approach for analyzing and managing macrofinancial risks” With the growing connectedness of global markets that has taken place over time, analyzing and managing macrofinancial risk has also increased in importance. Thus, the interactions between the household, corporate, financial and government sectors, as well as the corresponding risks, must also be understood, measured and monitored. In our paper (Merton et al., 2013), we use the Granger causality test in order to model and measure the degree of connectedness deriving from the feedback loops of explicit and implicit guarantees set between sovereigns, banks and insurers. Risks are evaluated through credit default swap (CDS) market data and a version of the Merton Model (based on Contingent Claims Analysis or CCA). The sample encompassed 17 sovereigns, 63 banks and 39 insurance companies. An analysis of the number of connections points to a much greater connectedness among financial institutions during and after the crisis periods with regard to all areas covered. The data suggest that the degree of connectedness across different types of entities changes over time, indicating that models capable of capturing these dynamics are needed in order to monitor the system. “Econometric measures of connectedness and systemic risk in the financial and insurance sectors” In this article (Billio et al., 2012), my coauthors and 6
Research • Center of Excellence SAFE • Quarter 1/2014 I targeted our research on banks, insurers, hedge funds and brokers operating in the U.S. market. We proposed several econometric measures of connectedness based on principal component analysis and Granger causality networks applied to the monthly returns of the target institutions. The aim of our research was to capture the network of causal relationships between the largest financial institutions across the four types of institutions mentioned. Moreover, we tried to understand whether these measures of connectedness are meaningful indicators of the condition of the financial markets and to discover, in the light of the last financial crisis, which are the key banks, insurers, hedge funds and brokers as far as systemic risk is concerned. The empirical analysis conducted shows that the degree of connectedness increases before and during crises, and the pattern of the last financial crisis suggests that banks are more central to systemic risk than the so-called “shadow banking system”. Moreover, the proposed measures are able to reflect periods of market dislocation and distress with promising out-of-sample characteristics. “Sovereign, bank and insurance credit spreads: connectedness and system networks” As banking and insurance system risk has become an important element in the determination of sovereign risk and vice-versa, in this article (Billio et al., Figure 1: Network diagrams of linear Granger-causality relationships that are statistically significant at the 1% level between banks (red), insurers (black), and sovereigns (blue) – from July 2004 to June 2007 (left) and from April 2009 to March 2012 (right) 2013) we focused on the relationship between the The application of the Granger causality test to ELR private financial sector and sovereign risk, aiming data allowed us to ascertain the evolution of connectedness. An empirical analysis showed that the to highlight the connectedness among financial institutions and sovereigns. To that extent, we applied several econometric measures of connect- in our sample is highly connected in a very dynamic system of banks, insurers and sovereigns included edness based on Granger causality networks to way. These interconnections are not symmetric. An changes in the sovereign risk of European countries analysis of the number of connections points to a and the credit risk of major European, U.S. and Japanese banks, broker-dealers and insurers. Financial stitutions after the crisis than before (see Figure 1). much greater connectedness among financial in- institution and sovereign risk is measured using Expected Loss Ratio (ELR) data from the Moody’s KMV and sovereigns were much more connected during Moreover, time series showed that banks, insurers database. This uses equity, equity volatility and the and after the global financial crisis than before. In default barrier (from accounting information) to conclusion, the proposed financial network measures can be utilized as early warning indicators get the “distance-to-distress” which it then maps to a default probability using a pool of 30 years of and to assess the complexity of the financial system. This framework can also be employed for default information. the analysis of shocks, spillovers and tradeoffs with regard to different policy alternatives. References Billio, M., Getmansky, M., Gray, D., Lo, A. W., Merton, R. C., Pelizzon, R. (2013) “Sovereign, Bank and Insurance Credit Spreads: Connectedness and System Networks”, MIT Working Paper. Billio, M., Getmansky, M., Gray, D., Lo, A. W., Pelizzon, R. (2012) “Econometric measures of connectedness and systemic risk in the financial and insurance sectors”, Journal of Financial Economics, Vol. 104, Issue 3, pp. 535-559. De Bandt, O., Hartmann, P. (2000) “Systemic Risk: A Survey”, European Central Bank Working Paper Series, No. 35. Merton, R. C., Billio, M., Getmansky, M., Gray, D., Lo, A. W., Pelizzon, R (2013) “On a New Approach for Analyzing and Managing Macrofinancial Risks”, Financial Analysts Journal, Vol. 69, Issue 2. Merton, R. C. (1974) “On the Pricing of Corporate Debt: The Risk Structure of Interest Rates”, Journal of Finance, Vol. 29, pp. 449-470. 7
Page 1 and 2: Newsletter at the House of Finance
Page 3 and 4: Editorial • Center of Excellence
Page 5: Research • Center of Excellence S
Page 9 and 10: Research • Center of Excellence S
Page 11 and 12: Policy • Center of Excellence SAF
Page 13 and 14: Selected Publications • Center of
Page 15 and 16: Events • Center of Excellence SAF

Newsletter

Create successful ePaper yourself

Delete template?

Save as template?